Systematic Review
org
Two-dose versus single-dose methotrexate for
treatment of ectopic pregnancy: a meta-analysis
Snigdha Alur-Gupta, MD; Laura G. Cooney, MD; Suneeta Senapati, MD, MSCE; Mary D. Sammel, ScD;
Kurt T. Barnhart, MD, MSCE
E ctopic pregnancies occur in approx-
imately 2% of pregnancies, yet ac-
count for 9% of maternal mortality and
are the leading cause of pregnancy-related
death in the first trimester.1 Advance-
ments in imaging technology and
protocols to screen women at risk have
led to earlier detection of ectopic
pregnancies.2e4 As more women with
ectopic pregnancies present as clinically
stable without concern for rupture, op-
tions for treatment have expanded beyond
surgical management to medical man-
agement. The mainstay of medical man-
agement has been methotrexate, a folate
antagonist that binds to dihydrofolate
reductase, leading to downstream inhibi-
tion of DNA synthesis and repair as well as
cell replication.3,5 Multiple publications
From the Division of Reproductive
Endocrinology and Infertility (Drs Alur-Gupta,
Cooney, Senapati, and Barnhart), University of
Pennsylvania, Philadelphia, PA; and Department
of Biostatistics and Epidemiology (Dr Sammel),
University of Pennsylvania, Philadelphia, PA.
Received Oct. 31, 2018; revised Dec. 18, 2018;
accepted Jan. 2, 2019.
Dr Cooney is currently at the Division of
Reproductive Endocrinology and Infertility,
University of Wisconsin, Madison, WI.
The authors report no conflict of interest.
Dr Barnhart is supported by R01 HD 036455.
Snigdha Alur-Gupta and Laura Cooney are
supported by the NIH T32 Training Grant:
HD007440. These sources did not play any role
in study design, in the analysis or interpretation
of data, in the writing of the report, or in the
decision to submit the article for publication.
Partially presented in poster form at the 73rd
Scientific Congress and Expo, American Society
for Reproductive Medicine, San Antonio, TX,
Oct. 28
Nov. 1, 2017.
Corresponding author: Snigdha Alur-Gupta,
MD. Snigdha.alur-gupta@uphs.upenn.edu
0002-9378/$36.00
ª 2019 Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.ajog.2019.01.002
ajog.
OBJECTIVE: To compare the treatment success and failure rates, as well as side effects
and surgery rates, between methotrexate protocols.
DATA SOURCES: PubMed, Embase, and the Cochrane library searched up to July 2018.
STUDY ELIGIBILITY CRITERIA: Randomized controlled trials that compared women with
ectopic pregnancies receiving the single-dose, two-dose, or multi-dose methotrexate
protocols.
STUDY APPRAISAL AND SYNTHESIS METHODS: Odds of treatment success, treatment
failure, side effects, and surgery for tubal rupture, as well as length of follow-up until
treatment success, were compared using random and fixed effects meta-analysis.
Sensitivity analyses compared treatment success in the groups with high human cho-
rionic gonadatropin (hCG) values and a large adnexal mass, as defined by individual
studies. The Cochrane Collaboration tool was used to assess risk of bias.
RESULTS: The 2-dose protocol was associated with higher treatment success
compared to the single-dose protocol (odds ratio [OR], 1.84; 95% CI, 1.13, 3.00). The
2-dose protocol was more successful in women with high hCG (OR, 3.23; 95% CI,
1.53, 6.84) and in women with a large adnexal mass (OR, 2.93; 95% CI, 1.23, 6.9).
The odds of surgery for tubal rupture were lower in the 2-dose protocol (OR, 0.65;
95% CI, 0.26, 1.63), but this was not statistically significant. The length of follow-up
was 7.9 days shorter for the 2-dose protocol (95% CI,
12.2,
3.5). The odds of
side effects were higher in the 2-dose protocol (OR, 1.53; 95% CI, 1.01, 2.30).
Compared to the single-dose protocol, the multi-dose protocol was associated with a
nonsignificant reduction in treatment failure (OR, 0.56; 95% CI, 0.28, 1.13) and a
higher chance of side effects (OR, 2.10; 95% CI, 1.24, 3.54). The odds of surgery for
tubal rupture (OR, 1.62; 95% CI, 0.41, 6.49) and time to follow-up (OR,
1.3; 95%
CI,
5.4, 2.7) were similar.
CONCLUSION: The 2-dose methotrexate protocol is superior to the single-dose protocol
for the treatment of ectopic pregnancy in terms of treatment success and time to suc-
cess. Importantly, these findings hold true in patients thought to be at a lower likelihood of
responding to medical management, such as those with higher hCGs and a large adnexal
mass.
Key words: doses, ectopic pregnancy, medical management, methotrexate, protocol,
tubal pregnancy
have demonstrated comparable efficacy of folinic acid rescue, alternating with
medical management to surgical man- methotrexate doses, to decrease side ef-
agement in the treatment of stable ectopic fects. The single-dose protocol was intro-
pregnancies.6-9 The most widely used duced to reduce the number of visits, but
methotrexate protocols include single- often requires additional treatment and
dose, 2-dose, and multi-dose protocols.3 follow-up. The 2-dose protocol was
The multi-dose protocol can be benefi- introduced with the goal of achieving a
cial for patients at higher risk for failing balance between the benefits of increased
medical management and thus requiring treatment success from additional doses of
additional doses.5,10 However, because of methotrexate, while using the same
its frequency of administration, the multi- convenient visit schedule as the single-
dose protocol requires the addition of dose protocol.11
MONTH 2019 American Journal of Obstetrics & Gynecology 1
Systematic Review ajog.org

Study selection
AJOG at a Glance Two independent reviewers (S.A.G. and
Why was this study conducted? L.C.) used the above-stated eligibility
This study was conducted to compare the odds of treatment success, side effects, criteria to screen all article titles and
surgery for ruptured ectopic pregnancy, and length of follow-up of commonly abstracts for inclusion. RCTs in human
used methotrexate protocols for the treatment of ectopic pregnancy. subjects with published manuscripts
were considered eligible. A flow diagram
Key findings of study selection is provided in Figure 1.
The 2-dose protocol was superior to the single-dose protocol in treatment suc-
cess, including in women at higher risk for failure, such as those with high human Data extraction
chorionic gonadatropin and large adnexal mass. Full texts of potentially eligible studies
were extracted and examined for the
What does this add to what is known? following data: year of study, number of
This adds an updated meta-analysis of a 2-dose versus a single-dose protocol and subjects, location of subject recruitment,
additional analyses of a multi-dose versus a single-dose protocol using only mean age of subjects, mean body mass
quality randomized controlled trials. index (BMI) of subjects, pretreatment
human chorionic gonadatropin (hCG)
values, pretreatment adnexal mass
The success rates of medical manage- incidence, and surgery rates among the diameter, description of methotrexate
ment of ectopic pregnancies have varied methotrexate protocols. protocols used, randomization and
with a range of 70e90% for the single- blinding processes, inclusion and exclu-
dose,12-14 80e90%14-16 for the 2-dose, Materials and Methods sion criteria, and definition of outcomes
and 89e96% for the multi-dose pro- Eligibility criteria measured, including treatment success,
tocols.12,13,17 Variation in rates may be The systematic review and meta- length of follow-up, side effects, and
influenced by the population being analysis were performed by strictly surgery for tubal rupture, as well as re-
studied, criteria for administering the following the Preferred Reporting sults for these outcomes. Several at-
medication, and definition of treatment Items for Systematic Reviews and tempts were made to electronically
success. Treatment failure rates, or Meta-Analyses (PRISMA) guidelines. contact authors of eligible studies that
chance of failure with a particular pro- The population being studied was did not explicitly contain the above in-
tocol, are also important to consider, as women with an ectopic pregnancy formation (Figure 1).
they can be clinically useful when coun- diagnosed by transvaginal ultrasound.
seling patients and can be a driver in Interventions included the single-dose, Assessment of risk of bias
deciding which protocol to recommend. 2-dose, or multi-dose methotrexate The Cochrane Collaboration tool was
Although several studies have protocol (Table 1). Comparisons used to assess the risk of bias. Two au-
attempted to compare one protocol to included the single-dose to the 2-dose thors (S.A.G. and L.C.) independently
another, they are limited by their retro- protocol, and the single-dose to the reviewed the included studies and
spective nature, nonstandard protocols, multi-dose protocol. Associations of assigned values of low, uncertain, or high
and heterogeneous definition of out- interest for binary outcomes are pre- risk to the 6 domains outlined in the tool
comes.18-21 Randomized controlled tri- sented as odds ratios (OR) with 95% (Figure 2).
als (RCTs) are considered the gold confidence intervals (CI). Only RCTs
standard when evaluating such ques- published as manuscripts with clear Data synthesis for meta-analysis
tions, because of their systematic, randomization protocols were included The following primary outcome mea-
reproducible approach with minimiza- in this analysis. sure was analyzed for included studies:
tion of confounding through the process treatment success (as reported in indi-
of randomization. Although there have Information sources and search vidual studies). Treatment failure
been other meta-analyses on this topic, strategy (defined as not achieving treatment
several do not include more recent Studies were identified by searching success with the stated protocol), which
studies, thus providing an opportunity PubMed, Embase, and the Cochrane li- is the weighted inverse of treatment
to update our understanding.2,9,12,22 brary in July 2018, with no starting date success and provides an alternative tool
Others are limited by inclusion of restrictions. Combinations of the to use when counseling patients, was also
retrospective studies and abstracts12,23 or following keywords were used to identify described. Secondary outcomes analyzed
lack of treatment failure rate report- the studies: “methotrexate”, “ectopic included side effects, surgery for tubal
ing.22,23 Our study sought to compare, pregnancy”, “tubal pregnancy”, “dose”, rupture, and length of follow-up in days
via a meta-analysis, quality RCTs evalu- and “protocol”. No filters were applied (as defined by individual studies). Side
ating the treatment success, side effect for language or location. effects reported in individual studies

2 American Journal of Obstetrics & Gynecology MONTH 2019

ajog.org Systematic Review

TABLE 1
Methotrexate protocols
Single-dose Two-dose Multi-dose
2 2
Day 1 Administer MTX 50 mg/m IM, obtain Administer MTX 50 mg/m IM, obtain Administer MTX 1 mg/kg IM, obtain
serum hCG serum hCG serum hCG
Day 2 — — —
Day 3 — — Administer second dose MTX 1 mg/kg
IM, obtain serum hCG; if >15% drop,
stop MTX and follow hCG levels weekly.
If <15% drop, proceed with plan
Day 4 Obtain serum hCG Administer second dose MTX 50 Administer leucovorin 0.1 mg/kg IM
mg/m2 IM
Day 5 — — Obtain serum hCG. If >15% drop, stop
MTX and follow hCG levels weekly. If
<15% drop, proceed with plan.
Administer third dose MTX 1 mg/kg IM
Day 6 — — Administer leucovorin 0.1 mg/kg IM
Day 7 Obtain serum hCG. If >15% drop, Obtain serum hCG. If >15% drop, Obtain serum hCG. If >15% drop, stop
follow hCG levels weekly. If <15% follow hCG levels weekly. If <15% MTX and follow hCG levels weekly. If
drop, administer second dose drop, administer third dose MTX <15% drop, proceed with plan.
MTX 50 mg/m2 IM 50 mg/m2 IM Administer fourth dose MTX 1 mg/kg IM
Day 8 — — Administer leucovorin 0.1 mg/kg IM
Day 11 — Obtain serum hCG. If >15% drop, —
follow hCG levels weekly. If <15%
drop, administer fourth dose MTX
50 mg/m2 IM
Day 14 Obtain serum hCG, if >15% drop, Obtain serum hCG, if >15% drop, Obtain serum hCG. If >15% drop, stop
follow hCG levels weekly. If <15% follow hCG levels weekly. If <15% MTX and follow hCG levels weekly. If
drop, administer third dose MTX 50 drop, consider surgery <15% drop, proceed with plan.
mg/m2 IM Administer fifth dose MTX 1 mg/kg IM
Day 21 Obtain serum hCG. If >15% drop, — Obtain serum hCG. If >15% drop, follow
follow hCG levels weekly. If <15% hCG levels weekly. If <15% drop,
drop, consider surgery consider surgery
hCG, human chorionic gonadatropin; IM, intramuscularly; MTX, methotrexate.
Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.

included nausea, diarrhea, mucositis,
abdominal pain, and laboratory test ab-
normalities. Fixed effects meta-analysis
was used to report odds ratios (OR)
with 95% confidence intervals (CI) for
the outcomes with low heterogeneity
including treatment success and failure,
surgery for tubal rupture, and side ef-
fects, whereas random effects meta-
analysis was used to report mean days
with 95% CI for the outcome of length of
follow-up (Figures 3 and 4).
Sensitivity analyses were conducted to
assess treatment success rates in high-
hCG groups (as reported in individual
studies, with a range of >3000e5500
mIU/mL for 2-dose vs single-dose pro-
tocols and >800 IU/L for multi-dose vs
single-dose protocols) and large adnexal
mass groups (as reported in individual
studies, with a range of >2e3.5 cm). The
heterogeneity among studies was evalu-
ated both via forest plots with 95% CI as
well as the I2 statistic, with P < .05
considered as statistically significant.
Publication bias was assessed via funnel
plots of the log OR (Supplementary
Figure 1). Analysis was conducted us-
ing STATA v14.2 (StataCorp, College
Station, TX).
Results
Study selection
The search strategy yielded 1013 results
in Embase, 408 in PubMed, and 116
in the Cochrane Library. The 521
duplicates were removed. The remaining
articles’ titles and abstracts were
screened, along with the bibliography
of recently published meta-analyses. A
total of 19 potentially eligible articles
were identified. Of these articles, 3
were excluded for lack of a published
manuscript,24-26 2 were excluded for
being retrospective studies,19,20 2 were
excluded for being prospective cohort
studies,21,27 2 were excluded because of
the inability to obtain the full text beyond
the abstract in English despite attempts
to contact the authors,28,29, 2 were
excluded for concerns regarding whether
and how randomization was per-
formed,30,31 and 1 article was excluded
for not using standard methotrexate

MONTH 2019 American Journal of Obstetrics & Gynecology 3

Systematic Review ajog.org

protocols.18 Therefore, 7 final publica-

FIGURE 1
tions were found to meet our inclusion
Flow diagram of study inclusion criteria.

Study characteristics
Study characteristics are presented in
Table 2. All studies were RCTs. Study
sizes ranged from 70 to 160 total pa-
tients. Although all studies reported rates
of treatment success, only half reported
side effects or length of follow-up. Five
reported rates of surgery for tubal
rupture: 3 in the single-dose vs 2-dose
group, and 2 in the single-dose vs
multi-dose group. Although the studies
by Saadati et al14 and Guvendang et al13
reported data on rates of surgery, indi-
cation for surgery as rupture or elective
surgery was not specified; therefore these
studies were not included for this
outcome.

Risk of bias of included studies

Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019. Results of bias assessment via the
Cochrane Collaboration tool are

FIGURE 2
Risk of bias assessment

Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.

4 American Journal of Obstetrics & Gynecology MONTH 2019

ajog.org Systematic Review

FIGURE 3
A, Forest plot: 2-dose vs single-dose treatment success. B, Forest plot: 2-dose vs single-dose treatment failure.
C, Forest plot: 2-dose vs single-dose Treatment success in high-HCG group (defined by individual studies with a
range of >3000e5500 mIU/mL). D, Forest plot: 2-dose vs single-dose Treatment success in large group (defined
by individual studies with a range of >2e3.5 cm). E, Forest plot: 2-dose vs single-dose side effects. F, forest plot:
2-dose vs single-dose surgery for ruptured ectopic pregnancy. G, Forest plot: 2-dose vs single-dose length of
follow-up

Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019. (continued)

presented in Figure 2. Although all performed. Although several studies outcome measures which are mainly
studies discussed random sequence were not explicit in reporting blinding to objective. The study by Saleh et al,32
generation clearly, some were not clear outcomes, reviewers believed that this however, was scored as having a high
as to how allocation concealment was would be unlikely to significantly skew degree of bias in this category because of

MONTH 2019 American Journal of Obstetrics & Gynecology 5

Systematic Review
FIGURE 3
Continued
Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.
the description of differential counseling
regarding elective surgery based upon
group.32 Blinding of personnel was
similarly not explicitly stated in multiple
studies. Although blinding of personnel
may not have been as practically feasible
because of the requirement for
6 American Journal of Obstetrics & Gynecology MONTH 2019
intramuscular injections, it is possible
that this could have affected patient
reporting of side effects, which was 1 of
the subjective outcomes measured. The
study by Saleh et al32 was scored high in
this category, as envelopes were opened
in front of patients, whereas the study by
ajog.org
(continued)
Guvendang et al13 was also scored high,
as they discussed that the study team was
not blinded. The study by Saadati et al14
hospitalized patients during treatment
and discharged them when hCG was less
than 200 mIU/mL; therefore, reporting
of outcomes may have been biased away
ajog.org
FIGURE 3
Continued
Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.
from the null because of a nonstandard
follow-up protocol as well as definition
of treatment success. Publication bias
was not noted to be significant when
looking at studies that compared either
the 2-dose to the single-dose protocol or
studies that compared the multi-dose to
the single-dose protocol (Supplementary
Figure 1).
Meta-analysis results
Single-dose vs 2-dose protocols. Meta-
analysis results are shown as forest plots
in Figure 3. For the primary outcome of
treatment success, 4 studies were iden-
tified comparing the single-dose to the
2-dose protocol, with the 2-dose proto-
col associated with 1.84 times the odds of
achieving treatment success (95% CI,
MONTH 2019 American Journal of Obstetrics & Gynecology
Systematic Review
(continued)
1.13, 3.00) compared to the single-dose
protocol14,15,32,33 (Figure 3a). Odds of
treatment failure were 0.54 times lower
in the 2-dose protocol (95% CI, 0.33,
0.89) (Figure 3b). For the secondary
outcome of side effects, 4 studies were
identified with a combined odds of side
effects that were 1.53 times higher in the
2-dose protocol compared to the single-
7
Systematic Review
FIGURE 3
Continued
Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.
dose protocol (95% CI, 1.01,
2.30).14,15,32,33 (Figure 3e). Odds of sur-
gery for tubal rupture were lower when
comparing the 2-dose to the single-dose
protocol (OR, 0.65; 95% CI, 0.26, 1.63),
but the difference was not statistically
significant (Figure 3f). The length of
follow up was 7.9 days shorter for the 2-
dose protocol compared to the single-
dose protocol (95% CI:
12.2,
3.5)
(Figure 3g).
Sensitivity analyses of 4 studies in the
high hCG groups revealed a 3.23 times
higher odds of treatment success with
the 2-dose protocol as compared to the
single-dose protocol (95% CI, 1.53,
6.84)14,15,32,33 (Figure 3c). Evaluation of
treatment success in the large adnexal
mass groups from 3 studies showed a
2.92 times higher odds of treatment
success with use of the 2-dose protocol as
compared to the single-dose protocol
(95% CI, 1.23, 6.93)14,15,32 (Figure 3d).
Single-dose vs multi-dose protocols. We
also identified 3 studies comparing
treatment success rates for single-dose vs
multi-dose protocol. The multi-dose
protocol is associated with a higher,
8 American Journal of Obstetrics & Gynecology MONTH 2019
albeit nonsignificant, odds of treatment
success compared to the single-dose
(OR, 1.79; 95% CI, 0.89, 3.62)
13,17,22,34 (not shown). Similarly, the
multi-dose was associated with a
nonsignificant 0.56 times lower odds of
treatment failure (95% CI, 0.28, 1.13)
(Figure 4a). The odds of side effects were
significantly higher in the multi-dose
protocol compared to the single-dose
protocol (OR, 2.10; 95% CI, 1.24,
3.54)13,17,34 (not shown). The odds of
surgery for tubal rupture and length of
follow-up were comparable between the
multi-dose and single-dose protocols
(OR, 1.62; 95% CI, 0.41, 6.49)
(Figure 4d) and
1.3 days (95%
CI,
5.4, 2.7), respectively (Figure 4e).
Only 1 study was identified assessing
treatment success when initial hCG was
high, when comparing the multi-dose to
the single-dose protocol, and found a
nonsignificant 2.00 times higher odds of
treatment success (95% CI, 0.54, 7.44)13
(Figure 4b). Only 1 study was also
identified assessing treatment success
with larger adnexal mass when
comparing the multi-dose to the single-
dose protocol, and found a continuation of treatment. Taken
ajog.org
nonsignificant 1.63 times higher odds of
treatment success (95% CI, 0.38, 6.96)13
(Figure 4c).
Comment
Overall, the 2-dose protocol was found
to result in a significantly higher odds of
treatment success, and thus a signifi-
cantly lower odds of treatment failure,
when compared to the single-dose pro-
tocol. These findings held true in pa-
tients with higher hCGs as well as in
patients with large adnexal mass as
defined by the individual studies. In
addition, the length of follow-up for
women receiving the 2-dose protocol
was more than 1 week shorter than for
women receiving the single-dose proto-
col. There was also a non
statistically
significant reduction in the odds of sur-
gery for tubal rupture with the use of the
2-dose protocol. The 2-dose protocol did
have a higher rate of side effects, but it
should be noted that most side effects
described in the included studies were
mild and transient. No patients required
hospitalization or long-term manage-
ment, nor did side effects preclude
ajog.org Systematic Review

FIGURE 4
A, Forest plot: multi-dose vs single-dose treatment failure. B, Forest plot: multi-dose vs single-dose treatment
success in high-hCG group (defined by study as >800 IU/L). C, Forest plot: multi-dose vs single-dose treatment
success in large size group (defined by study as >2 cm). D, Forest plot: multi-dose vs single-dose surgery for
ruptured ectopic pregnancy. E, Forest plot: multi-dose vs single-dose length of follow-up

Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019. (continued)

together, the 2-dose protocol is superior
to the single-dose protocol and should
be considered first-line therapy.
A meta-analysis has recently
compared treatment success and side
effects rates between multi-dose and
single-dose protocols.22 Given there
have been no additional RCTs evaluating
this comparison since this analysis was
performed, we validated these findings
and focused the analysis on odds of
failure, odds of surgery for tubal rupture,
length of follow-up, and sensitivity ana-
lyses, which were not performed in the
prior meta-analysis. Overall there is a
non
statistically significant trend to-
ward lower treatment failure and higher
treatment success, with use of the mul-
tidose protocol. In addition, we found
that the length of follow-up was only
roughly 1 day shorter in the multi-dose

MONTH 2019 American Journal of Obstetrics & Gynecology 9

Systematic Review
FIGURE 4
Continued
Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.
protocol and odds of surgery for tubal
rupture were not significantly different.
The quality of a meta-analysis is
directly dependent on the quality of the
studies included in the analysis. Previous
meta-analyses have been limited by
including retrospective or observational
studies.12 Moreover, additional ran-
domized trials have been conducted
10 American Journal of Obstetrics & Gynecology MONTH 2019
since the publication of recent analyses.
For example, the meta-analysis by Yang
et al22 did not include the study by Saleh
et al32 performed in 2016, and therefore
did not find statistically significant dif-
ferences when comparing the 2-dose to
the single-dose protocol. The meta-
analysis by Yuk et al.23 included poor-
quality data, including a meeting
ajog.org
(continued)
abstract without a full published manu-
script26 as well as RCTs that did not
specify how randomization was per-
formed or how many patients were in
each arm30 and that described their
randomization process as “patients were
alternatively selected”31 (and yet had
different numbers of patients in each
arm). Attempts were made to
ajog.org
FIGURE 4
Continued
Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.
electronically contact both authors
regarding these study details prior to
exclusion from our meta-analysis.
This meta-analysis has several
strengths, most importantly its rigid in-
clusion criteria. Only RCTs with clearly
stated methodology were included,
limiting the chances that results would be
biased by flaws in study design or
execution. In addition to treatment suc-
cess reporting, odds of failure were also
calculated, which can provide an addi-
tional useful tool and way to conceptu-
alize data when counseling patients.
The meta-analysis was limited by the
relatively few RCTs conducted on this
topic, particularly for the single versus
multi-dose protocols. In this category, 3
total quality RCTs were identified
assessing the main outcomes of treat-
ment success. It is possible that the
reason that our data cannot confirm a
higher success rate (and lower failure
rate) for the multi-dose protocol is lack
of power or inherent bias in the few
studies comparing multi-dose to single-
dose therapy. Only 1 of these studies
contained data for subgroup sensitivity
analyses. Based on the odds of treatment
success found with this meta-analysis,
we calculated that an additional study
with 70 patients per arm would be
needed for the odds of treatment success
to be significantly higher in the multi-
dose group as compared to the single-
dose group (assuming that the same
differences were to be found in 1
additional study). In addition, the meta-
analysis is limited in the ability to eval-
uate effectiveness in multiple subgroups.
For example, it is possible that there are
clinical situations when a single dose of
methotrexate is sufficient for the treat-
ment of an ectopic pregnancy, such as in
women with a low hCG value. However,
the limited data from the included ran-
domized clinical trials does not allow
such a conclusion from this meta-
analysis, and could be the focus of
future study.
Based on the pharmacokinetics of
methotrexate, it is logical that a second
dose would improve success rates
compared to a single dose, because a
second dose will affect a greater per-
centage of trophoblast cells in the S
phase (DNA synthesis). It is not clear
why an even greater number of doses, as
MONTH 2019 American Journal of Obstetrics & Gynecology
Systematic Review
part of the multi-dose protocol, does not
result in greater efficacy. This may be
related to low power, as stated earlier, or
to the possibility that the use of alter-
nating doses of leucovorin, in an attempt
to decrease side effects, may also limit
efficacy of the treatment. Multi-dose
treatment is not currently considered
first-line therapy and may be best
reserved for women with advanced
ectopic pregnancy or those in unusual
locations such as cervical, intestinal, or
ovarian ectopic pregnancy.35,36
In conclusion, the 2-dose protocol is
significantly superior to the single-dose
protocol in terms of odds of treatment
success and treatment failure. These
findings hold true in patients thought to
be at a lower likelihood of responding to
medical management, such as those with
higher hCGs and large adnexal mass.
Although the multi-dose protocol
showed similar trends when compared
to the single-dose protocol, none of these
parameters reached statistical signifi-
cance. Therefore, we would recommend
the 2-dose protocol as the first-line
protocol in patients being medically
managed for ectopic pregnancies. -
11
 Systematic Review
12 American Journal of Obstetrics & Gynecology MONTH 2019

TABLE 2
Included study characteristics
Total
Author /study Ectopic pregnancy patients Method of Definition of
location /year diagnostic criteria per arm Inclusion criteria Exclusion criteria randomization treatment success Other outcomes studied
Single versus Two dose
Song bHCG, TVUS, 46 46 1. bHCG <15,000 mIU/mL 1. Heterotopic pregnancy Randomly permuted bHCG <5 mIU/mL 1. Side effects
South Korea physical exam, 2. GS <4 cm or persistent tubal blocks with allocation 2. Length of follow-up
2015 Medical history 3. Hemodynamically stable pregnancy concealment (1:1 ratio) 3. Rate of operation
2. þFHR 4. Need for repeat doses
3. Suspected tubal rupture 3. Cost of care received
4. Laboratory test results 4. Days of work/school
contraindicating MTX use missed
5. Treatment satisfaction
Saadati bHCG, TVUS 38 38 1. bHCG <15,000 mIU/mL 1. Women with history Block randomization bHCG 200 mIU/mL 1. Side effects
Iran 2. Hemodynamically stable of liver, kidney disease, with enclosed envelopes 2. Length of follow-up
2015 or blood dyscrasia (1:1 ratio) 3. Rate of operation
2. Breastfeeding 4. Need for repeat doses
Hamed bHCG, TVUS, 78 79 1. bHCG of <15,000 mIU/ 1. Women suspected Computer-generated bHCG <15 mIU/mL 1. Side effects
Saudi Arabia progesterone and mL of having nonadnexal random numbers table within 2. Length of follow-up
2012 D&C when abortion 2. GS 4 cm ectopic pregnancy with opaque envelopes 6 weeks without 3. Rate of operation
suspected 3. Hemodynamically stable 2. Suspected tubal rupture surgery or 4. Need for repeat doses
4. Absence of FHR 3. Free fluid extending repeat dose
beyond Douglas pouch on
TVUS
4. Laboratory test results
contraindicating MTX use
Saleh bHCG, TVUS 80 80 1. bHCG 6000 mIU/mL 1. Hemodynamically Computer-generated bHCG <15 mIU/mL 1. Side effects
Egypt 2. GS 4 cm unstable randomization with within 6 weeks 2. Length of follow-up
2016 3. Hemodynamically stable 2. Suspected tubal rupture sealed, opaque without surgery or 3. Rate of operation
4. Absence of FHR 2. Uncertain diagnosis envelopes repeated dose 4. Need for repeat doses
5. <300 mL 3. Falling bHCGs
hemoperitoneum 4. Nonadnexal ectopic
on TVUS pregnancy
5. Laboratory test results
contraindicating MTX use
6. Breastfeeding
7. Immunodeficiency or use
of corticosteroids
Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019. (continued)

ajog.org
ajog.org Systematic Review

REFERENCES
1. Marion LL, Meeks GR. Ectopic pregnancy:

5. Outcomes of subsequent
Other outcomes studied
History, incidence, epidemiology, and risk

3. Need for repeat doses

4. Need for repeat doses

4. Need for repeat doses

2. Length of follow-up

2. Length of follow-up

2. Length of follow-up
factors. Clin Obstet Gynecol 2012;55:

3. Rate of operation

3. Rate of operation
376–86.
2. Hajenius PJ, Mol F, Mol BW, Bossuyt PM,

1. Side-effects
1. Side effects

1. Side effects
Ankum WM, van der Veen F. Interventions for

pregnancies
tubal ectopic pregnancy. Cochrane Database
Syst Rev 2007:CD00324.
3. ACOG. ACOG practice bulletin no. 193: tubal
ectopic pregnancy. Obstet Gynecol 2018;131:
e91–103.
treatment success

bHCG <15 mIU/mL

bHCG <15mIU/mL
4. Barnhart KT. Clinical practice. Ectopic preg-
bHCG <5 mIU/mL

within 6 weeks

within 6 weeks
nancy. N Engl J Med 2009;361:379–87.
Definition of

5. Barnhart K, Coutifaris C, Esposito M. The

pharmacology of methotrexate. Expert Opin
Pharmacother 2001;2:409–17.
6. Hajenius PJ, Engelsbel S, Mol BW, et al.
Randomised trial of systemic methotrexate
versus laparoscopic salpingostomy in tubal
Computer-generated block

Computer-generated block
randomization with sealed

randomization with sealed

randomization with sealed

pregnancy. Lancet 1997;350:774–9.

7. Kazandi M, Turan V. Ectopic pregnancy; risk
Computer-assisted

factors and comparison of intervention success

randomization

bHCG, beta human chorionic gonadotropin; D&C, dilation and curettage; FHR, fetal heart rate; GS, gestational sac; MTX, methotrexate; TVUS, transvaginal ultrasound. rates in tubal ectopic pregnancy. Clin Exp
Obstet Gynecol 2011;38:67–70.
Method of

envelopes

envelopes

envelopes

8. Krag Moeller LB, Moeller C, Thomsen SG,

et al. Success and spontaneous pregnancy
rates following systemic methotrexate versus
laparoscopic surgery for tubal pregnancies: a
3. Hepatic or renal disease

randomized trial. Acta Obstet Gynecol Scand

2009;88:1331–7.
1. Prior tubal surgery

9. Mol F, Mol BW, Ankum WM, van der Veen F,

2. Hemodynamically
Exclusion criteria

Hajenius PJ. Current evidence on surgery, sys-

temic methotrexate and expectant management
1. bHCG <15,000 mIU/mL None listed

1. bHCG <15,000 mIU/mL None listed

in the treatment of tubal ectopic pregnancy: a

unstable

systematic review and meta-analysis. Hum

Reprod Update 2008;14:309–19.
10. Lipscomb GH, McCord ML, Stovall TG,
1. bHCG reaching plateau or

Huff G, Portera SG, Ling FW. Predictors of

3. Hemodynamically stable

3. Hemodynamically stable

3. Hemodynamically stable
2. Adnexal mass <3.5 cm
increased by 50% in 48

2. Adnexal mass 3.5cm

5. Absence of bleeding in

success of methotrexate treatment in women

2. Adnexal mass 4cm

laparoscopic surgery or

with tubal ectopic pregnancies. N Engl J Med

4. Absence of FHR

4. Absence of FHR

1999;341:1974–8.
Inclusion criteria

11. Barnhart K, Hummel AC, Sammel MD,

Menon S, Jain J, Chakhtoura N. Use of "2-dose"
Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.

regimen of methotrexate to treat ectopic preg-

nancy. Fertil Steril 2007;87:250–6.
hours

TVUS

12. Barnhart KT, Gosman G, Ashby R,

Sammel M. The medical management of ectopic
pregnancy: a meta-analysis comparing "single
Included study characteristics (continued)

dose" and "multidose" regimens. Obstet Gyne-

patients
per arm

62 58

54 54

35 35

col 2003;101:778–84.
Total

13. Guvendag Guven ES, Dilbaz S, Dilbaz B,

Aykan Yildirim B, Akdag D, Haberal A. Com-
parison of single and multiple dose methotrexate
TVUS, Laparoscopic
Ectopic pregnancy
diagnostic criteria

therapy for unruptured tubal ectopic pregnancy:

a prospective randomized study. Acta Obstet
progesterone
bHCG, TVUS,

bHCG, TVUS

Gynecol Scand 2010;89:889–95.

14. Saadati N, Najafian M, Masihi S, Safiary S,
Single-dose vs multi-dose

surgery

Abedi P. Comparison of two different protocols

of methotrexate therapy in medical management
of ectopic pregnancy. Iranian Red Crescent Med
location /year
Author /study

J 2015;17:e20147.
15. Hamed HO, Ahmed SR, Alghasham AA.
Guvendang

Tabatabaii
Alleyassin
TABLE 2

Comparison of double- and single-dose meth-

Turkey
Guven

otrexate protocols for treatment of ectopic

2010

2005

2012
Iran

Iran

pregnancy. Int J Gynaecol Obstet 2012;116:

67–71.

MONTH 2019 American Journal of Obstetrics & Gynecology 13

Systematic Review
16. Gungorduk K, Asicioglu O, Yildirim G,
Gungorduk OC, Besimoglu B, Ark C. Compari-
son of single-dose and two-dose methotrexate
protocols for the treatment of unruptured
ectopic pregnancy. J Obstet Gynaecol 2011;31:
330–4.
17. Tabatabaii Bafghi A, Zaretezerjani F,
Sekhavat L, Dehghani Firouzabadi R,
Ramazankhani Z. Fertility outcome after treat-
ment of unruptured ectopic pregnancy with two
different methotrexate protocols. Int J Fertil Steril
2012;6:189–94.
18. Chen YX, Mao YY, Xie X. Efficacy and side
effects of methotrexate in the treatment of
ectopic pregnancy. Zhonghua Fu Chan Ke Za
Zhi 2003;38:749–51.
19. Lipscomb GH, Givens VM, Meyer NL,
Bran D. Comparison of multidose and single-
dose methotrexate protocols for the treatment
of ectopic pregnancy. Am J Obstet Gynecol
2005;192:1844–7; discussion 47-8.
20. Kim J, Jung YM, Lee DY, Jee BC. Pre-
treatment serum human chorionic gonadotropin
cutoff value for medical treatment success with
single-dose and multi-dose regimen of metho-
trexate in tubal ectopic pregnancy. Obstet
Gynecol Science 2017;60:79–86.
21. Mergenthal MC, Senapati S, Zee J, et al.
Medical management of ectopic pregnancy with
single-dose and 2-dose methotrexate pro-
tocols: human chorionic gonadotropin trends
and patient outcomes. Am J Obstet Gynecol
2016;215:590.
22. Yang C, Cai J, Geng Y, Gao Y. Multiple-
dose and double-dose versus single-dose
administration of methotrexate for the treat-
ment of ectopic pregnancy: a systematic review
14 American Journal of Obstetrics & Gynecology MONTH 2019
and meta-analysis. Reprod Biomed Online
2017;34:383–91.
23. Yuk JS, Lee JH, Park WI, Ahn HS, Kim HJ.
Systematic review and meta-analysis of single-
dose and non-single-dose methotrexate pro-
tocols in the treatment of ectopic pregnancy. Int
J Gynaecol Obstet 2018;141:295–303.
24. Cardamakis E, Kourounis G, Tzeferakos A,
Korantzis A, Ginopoulos P. Three different regi-
mens of methotrexate for the treatment of
ectopic pregnancy. XV FIGO World Congress of
Obstetrics & Gynecology 1997:92.
25. Shahrzad M. Comparison of single and
multiple-dose methotrexate therapy for ectopic
pregnancy: a clinical trial. Clinicaltrials.gov.
Study ID: NCT01662167. Available at: https://
clinicaltrials.gov/show/nct01662167. Accessed
July 16, 2018.
26. Klauser CK, May WL, Johnson VK,
Cowan BD, Hines RS. Methotrexate for ectopic
pregnancy: a randomized “single dose”
compared with “multidose” trial. Obstet Gynecol
2005;105:64S.
27. Olofsson JI, Poromaa IS, Ottander U,
Kjellberg L, Damber MG. Clinical and preg-
nancy outcome following ectopic pregnancy;
a prospective study comparing expectancy,
surgery and systemic methotrexate treat-
ment. Acta Obstet Gynecol Scand 2001;80:
744–9.
28. Golmohammadlou S, Hajishafiha M,
Hasankhani Z. Comparison of single and two
dose of methotrexate in treatment of ectopic
pregnancy. Iran J Reprod Med 2012;10:74.
29. Fakheri T, Nankali A, Shahlazadeh H,
Ataee M, Shahri MN. Comparison of single-dose
and multiple-dose systemic methotrexate
ajog.org
regimen for ectopic pregnancy. J Isfahan Med
Sch 2014;32:1054–60.
30. Zargar M, Razi T, Barati M. Comparision of
single and multidose of methotrexate in medical
treatment of ectopic pregnancy. Pakistan J Med
Sci 2008;24:586–9.
31. Amirian M, Rajaee M, Moayedi U,
Mohammadi F, Hosseini S. Comparison of sin-
gle and multiple-dose methotrexate therapy for
ectopic pregnancy: a clinical trial. Life Sci J
2013;10:564–7.
32. Saleh H. Double versus single dose meth-
otrexate regimens in management of undis-
turbed ectopic pregnancy. Crit Care Obstet
Gynecol 2016;2:6.
33. Song T, Kim MK, Kim ML, Jung YW,
Yun BS, Seong SJ. Single-dose versus two-
dose administration of methotrexate for the
treatment of ectopic pregnancy: a random-
ized controlled trial. Hum Reprod 2016;31:
332–8.
34. Alleyassin A, Khademi A, Aghahosseini M,
Safdarian L, Badenoosh B, Hamed EA. Com-
parison of success rates in the medical man-
agement of ectopic pregnancy with single-dose
and multiple-dose administration of metho-
trexate: a prospective, randomized clinical trial.
Fertil Steril 2006;85:1661–6.
35. Ramkrishna J, Kan GR, Reidy KL, Ang WC,
Palma-Dias R. Comparison of management reg-
imens following ultrasound diagnosis of nontubal
ectopic pregnancies: a retrospective cohort
study. Br J Obstet Gynecol 2018;125:567–75.
36. Hunt SP, Talmor A, Vollenhoven B. Man-
agement of non-tubal ectopic pregnancies at a
large tertiary hospital. Reprod Biomed Online
2016;33:79–84.
ajog.org Systematic Review

SUPPLEMENTARY FIGURE 1A
Funnel plot of publication bias: Two dose versus single dose protocols.
Points refer to individual study results. Dotted lines refer to 95%
confidence interval. The symmetric nature of the dots about the solid
vertical line (representing the average association measure) indicates
lack of publication bias

Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.

MONTH 2019 American Journal of Obstetrics & Gynecology 14.e1

Systematic Review ajog.org

SUPPLEMENTARY FIGURE 1B
Funnel plot of publication bias: Multi-dose versus single dose protocols.
Points refer to individual study results. Dotted lines refer to 95%
confidence interval. The symmetric nature of the dots about the solid
vertical line (representing the average association measure) indicates
lack of publication bias

Alur-Gupta. Meta-analysis of methotrexate protocols. Am J Obstet Gynecol 2019.

14.e2 American Journal of Obstetrics & Gynecology MONTH 2019